To address common challenges faced by internet hospitals—such as high patient volumes in key specialties including liver surgery, cardiology, and gastroenterology, along with the heavy burden of manual responses from physicians—the Laboratory of Natural Language Processing and Big Data Mining at East China University of Science and Technology, in collaboration with HealthCloud (Shanghai) Digital Technology Co., Ltd. and Renji Hospital, Shanghai Jiao Tong University School of Medicine, has developed multimodal data access and governance technologies, trained diagnostic and therapeutic large language models, and mined clinical pathways. The AI Doctor Assistant (AIDA) is a multimodal, specialty-specific large language model and intelligent decision-making platform built upon over 14 million service encounters from the internet hospital of Renji Hospital. It deeply integrates real-world clinical dialogue and data alongside multidisciplinary expert knowledge. Currently, AIDA has been deployed across 18 specialties at Renji Hospital, including liver surgery, cardiology, reproductive medicine, and rheumatology, serving tens of millions of residents in Shanghai. It has received positive feedback and generated substantial social benefits, significantly improving clinical efficiency while standardizing and digitizing the diagnostic and treatment capabilities of a leading hospital. This achievement has been selected for the first batch of high-quality dataset pilot application scenarios announced by the Shanghai Data Bureau and was showcased at the 2026 Dubai World Health Expo, demonstrating Shanghai’s innovation in the healthcare sector to the world. During the translation of research outcomes into practice, two papers have been accepted at ACL 2026, over ten patents have been filed, and five patents have been granted.
